The honest answer: partly, and it depends what the job actually is
A media buyer's job is really two separate jobs stacked on top of each other. One is repetitive execution: bid adjustments, budget pacing, pulling reports, pausing dead ad sets. The other is judgment: deciding what creative angle to test next, reading why a campaign actually failed, and knowing when a metric that looks bad is actually fine.
What AI already does better than a human
Checking account performance every few hours instead of once a week. Catching a budget drift or a targeting mistake within hours instead of days. Cross-referencing what the ad platform reports against what actually closed in your CRM, something most human media buyers never do consistently because it is tedious. This is grunt work at speed, and AI is simply faster and more consistent at it than a person checking a dashboard between other tasks. If you are still deciding what to spend in the first place, the free Facebook ads budget calculator works out a daily budget from your cost per lead and close rate.
What still needs a human
Deciding on a genuinely new creative angle. Reading a client's business model well enough to know which metric actually matters to them. Handling the conversation when something goes wrong. AI can flag that a campaign has been running the same angle for three months and is fatiguing. It cannot yet write the replacement angle from scratch and know it will land with a specific audience.
Where this is heading
The realistic setup right now is AI handling execution and monitoring around the clock, with a human making the judgment calls it surfaces. That is the model we run at KenjiAI, agents doing the repeatable work, a person deciding what to do about what they find.